A Design Environment for the Cost Optimization of Wireless Sensor Networks
University of Berlin, DE
Keywords: design environmen, cost optimization, wireless sensor networks
Abstract:Visions like Ambient Intelligence and Pervasive Computing expect for the future that a multitude of wireless sensor nodes will be integrated into everyday objects. A broad range of applications can only reach a large market volume if a very cost-efficient fabrication of these devices becomes realistic. Wireless sensor networks exhibit a lot of design trade-offs regarding system behaviour, hardware architecture and fabrication technology. A new approach for the hardware design of sensor networks will be presented based on a multi-criteria optimization by means of evolutionary algorithms. Thereby, the focus lies on the determination of the cost-driving components by using cost models. During an automated selection procedure of several implementation variants, not only expenditures for development and fabrication are considered, but also additional objectives like time to market or process complexity. A design environment was implemented in order to support the development of very cost-efficient sensor networks. The integration of specific cost models allows the designer to identify the appropriate architecture variants and technology options. The verification of the design tool resulted in valuable knowledge about sensor nodes especially for specific application scenarios. Prototypes of different architectures were designed and discussed regarding several options for cost optimization.